Other Automakers Cheating on Diesel? Not So Fast.

“No hard evidence,” though NOx numbers do exceed standards.

by on Oct.09, 2015

Diesels are inherently dirty, though modern pollution control systems can clean them up - if used properly.

In the wake of the VW diesel emissions scandal, a series of reports have suggested that other manufacturers might also be cheating, rigging their vehicles to do well on government tests while producing far more pollution in the real world.

The latest is a report in British newspaper The Guardian, its headline blaring, “Wide range of cars emit more pollution in realistic driving tests, data shows.” The paper cites real world test results from a UK lab that supposedly show some European diesel models emit up to 20 times the smog-causing oxides of nitrogen as EU rules permit.

The Last Word!

But that’s no surprise, and not necessarily a problem – never mind anything illegal — stressed Nick Molden, the CEO of that lab, Emissions Analytics, in an exclusive interview with TheDetroitBureau.com and NBCNews.

“They picked up the data…which intonate four other manufacturers have been conducting illegal activities,” said Molden, but in reality, “no one has any hard evidence of anyone doing anything clearly illegal.”

(Just following orders? Who was really responsible for VW diesel cheating? Click Here for more.)

The problem, Molden and other experts point out, is that government emissions tests are conducted under tightly controlled lab conditions that generally don’t reflect the way vehicles are driven in the real world. The European tests, in particular, don’t put a vehicle through the aggressive acceleration, uphill climbs and cold weather conditions a motorists is likely to experience.

“You’ve got a very gentle test cycle with a lot of loopholes in it,” said Molden.

In reality, most diesel vehicles produce almost no NOx when the engine and emissions system have been warmed up and they’re cruising smoothly down a flat highway. But that will spike under more aggressive driving, at times – briefly — to up to 20 times the supposed limit.

As a result, tests of about 200 diesel models have found that while the latest European standard permits the emission of 80 milligrams of NOx per kilometer, “on the road it will be four times the limit, or 320 mg,” said Molden, “and all legal.”

That position is echoed by Arvind Thiruvengadam, the research professor at West Virginia University who helped uncover the Volkswagen cheating.

“There would be differences between a certification test and a real world test,” emphasized Prof. Thiruvengadam, who added that, “I would be happy if a pass car is putting out just five to six times the standard. But that would only be under extreme conditions. And cruising down the highway, I would expect to see it producing much less than the standard.”

Diesel engines, both researchers noted, are particularly vulnerable to emissions spikes, especially when it comes to oxides of nitrogen. That has to do with the basic mechanics of the technology, a downside of the fact that diesels tend to deliver more stump-pulling torque, while also getting better mileage than comparable gasoline engines.

(VW says it will fix diesel emissions problem. Click Here for more on its plan.)

The latest diesels – which use turbochargers, high-pressure injectors and urea injection systems – have seen major reductions in emissions, especially NOx and sooty particulates. In fact, British researcher Molden says five of the newest models, which he declined to identify, are matching lab and real-world emissions numbers.

Those more advanced emissions systems are just rolling out in Europe but are mandated here in the U.S. But even then, Prof Thiruvengadam, of WVU, said short emissions spikes will likely be the norm for diesels. What matters most, he said, is that, on the whole, they are much cleaner than ever.

Both Thiruvengadam and Molden would like to see changes to emissions tests to make them better reflect real driving. They also think that revised testing would make it more difficult for a manufacturer, such as Volkswagen, that did try to cheat.

(VW planning to cut diesel line-up in U.S. For more, Click Here.)

Tags: , , , , , , , , , , , , , ,

8 Responses to “Other Automakers Cheating on Diesel? Not So Fast.”

  1. Warren Brown says:

    Nicely done, Paul. I hope this piece gets wide circulation. Thanks.–Warren

  2. GT101 says:

    FWIW- The Guardian is noted by many people for it’s sensational and inaccurate reporting. The half-truths in the story are intended to misrepresent and distort the public’s perception of clean diesel engine operation, due to technical ignorance and/or for financial gain. Sensationalism generates great revenues.

    As previously noted the VW diesels and likely all U.S. sold diesel vehicles meet the U.S. EPA dynamometer diesel emissions test requirements. Also as previously noted and confirmed by Mr. Molden, the real world emissions vs. the EPA or similar dynamometer test emissions are going to vary, just as the EPA fuel economy and actual driver mpg vary as much as 30%. What we do know is that all of the latest clean diesels are dramatically lower in all exhaust emissions including NOx. Some Europeans want to lower NOx similar to CA’s extreme requirement. Other folks believe the current engines are plenty clean.

    Mr. Molden’s statements regarding EPA test emissions results vs. actual driving results in fully compliant clean diesel engines is the same info. that other Euro auto manufacturer engineers have explained to me. This very important point however is missed by many people including some in the media who fail to comprehend and report the difference in emissions is not a violation of law or even a significant environmental pollutant because of the extremely low required emissions on clean diesels.

    • Paul A. Eisenstein says:

      GT, There are three levels of “compliance” with government tests:

      -Those vehicles which comply on the dyno and then meet the standards in real-world situations, as well. According to Molden, only five of the 200+ they’ve tested did so, all using the newest diesel emissions technologies;

      -There are those vehicles that comply on the dyno but run higher in real-world conditions. That is a result of several factors, including what was, Molden noted, testing that itself didn’t represent real-world conditions. In what I shall call Level II compliance (my term), makers may tune their hardware and software to maximum benefit for running on the dyno tests, but other than, say, such things as transmissions that shift at optimized points, there is nothing illegal here, and the things they tune will behave the same way in the real world if a driver operates the vehicle similarly to what happens on the dyno;

      -Then there is VW; the maker intentionally created software that ONLY performs to mandated levels on the dyno…and only when it RECOGNIZES it is on the dyno. That means that it was designed to sense such things as the steady barometric pressures used in a test cell, the lack of any steering input, etc. That qualifies, unabashedly, as cheating, as the vehicle will never comply on the real world because it was programmed to do so ONLY in the lab. Turn the steering wheel, adjust the pressure in the cell, etc., and it suddenly thinks it is on the road and disables some of the hardware, alters such things as shift timing, changes the timing of the ignition, etc….whatever it takes to deliver the performance and mileage targets VW set at the expense of emissions.

      As to my “Level II” compliance, both govt and makers are in a wink-wink, nudge-nudge sort of pattern here. The tests often don’t reflect what drivers really do and, arguably, neither do the standards. (This is particularly the case in Europe, per Mr. Molden.) Note that the EPA has a NTE, or Not-to-Exceed, clause in its emissions standards for heavy trucks. That means that for brief spurts — say charging up a hill with a full load in a dump truck — a heavy can exceed the emissions standard by x.x%. That is typically 1.5 TIMES the standard, but the EPA will negotiate with makers to allow variances reflecting impact on performance, mileage, durability, etc.

      Something along the lines of the NTE for real-world automotive testing is probably called for. A point to be made by both of my sources on this story: while diesel passenger cars also will exceed limits under hard acceleration or load (that aggressive hill climb), they just as typically produce almost NO measurable NOx when cruising down a freeway, for example. The point is balancing things out on an overall mg/mile level.

      One sidebar point: unfortunately, those spikes in NOx emissions often happen in urban environments where their impact on human health can be most severe, ie in stop-and-go city traffic, so simply allowing a diesel to spike emissions in town, then operate cleanly out in the country is not a truly balanced scenario.

      Paul A. Eisenstein
      Publisher, TheDetroitBureau.com

  3. GT101 says:


    All EPA and other dyno tests of the VW diesel powered autos showed that they passed the required emissions. The point being the engines can and do meet all emissions requirements when the controls are operated appropriately via proper ECU software.

    VW using sensor inputs to guess that an emissions test was being conducted is illegal. That isn’t disputed. The emissions device control duty cycles however do not need to run at 100% all of the time in normal vehicle operation to meet many emissions laws.

    The EPA and other regulators have confirmed these VW diesel engines do meet all emissions requirements when run on the EPA mandated dynamometer tests. The EPA has made no accusations of cheating on the dynamometer tests. The EPA has stated that the emissions results in the VW diesel dyno testing were fully compliant.

    The two emissions software control violations are the test detection software code that everyone agrees is illegal and the normal vehicle operation that uses lower emissions control device duty cycles than required to meet the emissions regs, 100% of the time. These are the two areas that VW must fix in their software according to the EPA. Tuning the emissions control duty cycles appropriately via the software to meet all EPA requirements is what VW needs to do and should have done from the beginning. Any additional hardware VW choses to add is to engine improve engine performance not to meet emissions.

    It would certainly appear that better designed and articulated emissions regulations are in order to prevent cheating and misunderstandings. It would also be useful if the diesel emissions regulations were based primarily on science and practicality and not politically motivated to deter the sales of clean diesels in order to promote sales of EVs that are quite impractical for most of society. When the U.S. has the lowest volume of diesel powered autos and the most stringent diesel emissions requirements in the world, it shows a clear political bias against clean diesel engines. There is no perfect solution to environmental air pollution issues and clean diesels should not be singled out by a technically illiterate administration to serve their personal agendas that do not concur with the public’s best interests.

    • Paul A. Eisenstein says:


      You keep sidestepping the issue: they meet the standards ONLY when they cheat. In other words, they are designed to detect when they are being subjected to very specifics test conditions, enable full emissions controls, and pass. But under such conditions they do NOT appear to deliver either the mileage and/or performance claimed for the same vehicles. Thus, if the conditions under which the vehicles “pass” is fraudulently achieved, meaning the powertrain is not in a normal operating mode, then it is not legally a pass. And under no real-world conditions can or will the VW engines actually meet the standards. This is NOT the same as, say, a Ford or a Chrysler product that has been optimized for the test but which could also deliver the same numbers on the road when driven in a similar manner to what the test attempts to simulate.
      This is, in effect, sending a ringer in to take a student’s final exam. In this case, the ringer and the real student are technically one and the same, but the software cheat operates only under the controlled circumstances it is designed to detect and nowhere else. If it DID work ON the road it would result in a loss of performance and/or fuel economy, as noted, and thus VW would be dinged in the separate EPA mileage tests, and consumers would be screaming about reduced performance and mileage.

      The bottom line is, if it requires a cheat to achieve the numbers, it is not legally passing. The tests require that the setup of the vehicle be what the automaker will use on the road, not something unique for the test.

      All else is irrelevant. They cheated. The results are invalid.

      Paul E.

      • GT101 says:

        Paul, no I am not sidestepping any issue. There are two distinctly different issues. One is the illegal detection of an emissions test. Two is the actual emissions test itself. The vehicles meet all required emissions when the EPA testing is conducted. Claiming that VW cheated on the tests implies the results are not real when they are. The proper way to report technical issues like this is the way the EPA website reports them so that consumers and owners know the specifics and are not misled into believing the vehicles won’t meet emissions, which they will with just a software upgrade.

        The bottom line is the VW diesel engines meet all emissions when dyno tested according to emissions regs. The problem is VW needs to fix their illegal and improper software to comply in normal driving. Implying that the engines don’t meet emissions by claiming VW cheated is technically incorrect and disingenuous, when you know differently.

        We’re going to have to agree to disagree because your POV is not supported by the EPA online documents. No where has the EPA claimed that VW’s engines do not pass the EPA requirements when run on the EPA dynamometer test cycle.

        • Paul A. Eisenstein says:

          GT, to wrap this up, the diesels MAY pass when tested, but the regs ALSO require that manufacturers do NOT specifically tune vehicles solely for the test process, which is what VW did. Even the VW insiders I talk to acknowledge that. Again, per my earlier notes, if a vehicle does well on the test but not as well on the road, that’s one thing. But to engineer a cheat specifically operational ONLY during testing is NOT legal. The same would be true for mileage and crash testing. A maker couldn’t for example, provide a modified vehicle that would score well in a crash test but then not market the same construction on the road.

          Paul E.

  4. GT101 says:

    I have already acknowledged many times that VW used illegal test detection code. That was never in dispute. What I take exception to is the improper use of the term cheating that implies the engines will not meet emissions when tested on the EPA dyno testing, which they clearly do. The EPA has not stated the engines do not comply with all regs in the EPA dyno testing but your stories like many in the media do not clearly explain the actual facts and as such mislead the public. That’s why it’s more appropriate to use the language the EPA uses because they even state that the engines do initially meet all regs in normal driving but over time the emissions control duty cycles are reduced to the point where the engines start to emit more exhaust pollution than allowed. No one would know the truth about the VW diesel engine emissions based on the reports by many media reports due to poor research or lack of understanding by many in the media.